trainer_state.json

{
  "best_metric": 0.5820802450180054,
  "best_model_checkpoint": "bert_base_lda_5_v1_mrpc/checkpoint-30",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 105,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.2099486589431763,
      "learning_rate": 4.9e-05,
      "loss": 0.6475,
      "step": 15
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6985294117647058,
      "eval_combined_score": 0.7542106100554595,
      "eval_f1": 0.8098918083462132,
      "eval_loss": 0.5918997526168823,
      "eval_runtime": 0.2723,
      "eval_samples_per_second": 1498.395,
      "eval_steps_per_second": 7.345,
      "step": 15
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.662345290184021,
      "learning_rate": 4.8e-05,
      "loss": 0.5596,
      "step": 30
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7107843137254902,
      "eval_combined_score": 0.7573855123444727,
      "eval_f1": 0.8039867109634552,
      "eval_loss": 0.5820802450180054,
      "eval_runtime": 0.274,
      "eval_samples_per_second": 1488.852,
      "eval_steps_per_second": 7.298,
      "step": 30
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5133867263793945,
      "learning_rate": 4.7e-05,
      "loss": 0.4573,
      "step": 45
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6593137254901961,
      "eval_combined_score": 0.7070819068368088,
      "eval_f1": 0.7548500881834215,
      "eval_loss": 0.6319071650505066,
      "eval_runtime": 0.3008,
      "eval_samples_per_second": 1356.234,
      "eval_steps_per_second": 6.648,
      "step": 45
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.9438729286193848,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3282,
      "step": 60
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6299019607843137,
      "eval_combined_score": 0.6727664229533621,
      "eval_f1": 0.7156308851224105,
      "eval_loss": 0.8020564317703247,
      "eval_runtime": 0.2743,
      "eval_samples_per_second": 1487.55,
      "eval_steps_per_second": 7.292,
      "step": 60
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.2352547645568848,
      "learning_rate": 4.5e-05,
      "loss": 0.1884,
      "step": 75
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6470588235294118,
      "eval_combined_score": 0.6993914807302232,
      "eval_f1": 0.7517241379310345,
      "eval_loss": 0.9919410943984985,
      "eval_runtime": 0.2747,
      "eval_samples_per_second": 1485.427,
      "eval_steps_per_second": 7.282,
      "step": 75
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.532591342926025,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0919,
      "step": 90
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6740196078431373,
      "eval_combined_score": 0.7209539574992301,
      "eval_f1": 0.767888307155323,
      "eval_loss": 1.1951086521148682,
      "eval_runtime": 0.2851,
      "eval_samples_per_second": 1430.86,
      "eval_steps_per_second": 7.014,
      "step": 90
    },
    {
      "epoch": 7.0,
      "grad_norm": 5.2938971519470215,
      "learning_rate": 4.3e-05,
      "loss": 0.0649,
      "step": 105
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6568627450980392,
      "eval_combined_score": 0.7081564584596725,
      "eval_f1": 0.7594501718213058,
      "eval_loss": 1.3058247566223145,
      "eval_runtime": 0.2783,
      "eval_samples_per_second": 1466.056,
      "eval_steps_per_second": 7.187,
      "step": 105
    },
    {
      "epoch": 7.0,
      "step": 105,
      "total_flos": 3377819728711680.0,
      "train_loss": 0.3339719607716515,
      "train_runtime": 62.0964,
      "train_samples_per_second": 2953.472,
      "train_steps_per_second": 12.078
    }
  ],
  "logging_steps": 1,
  "max_steps": 750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3377819728711680.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}